Objective There is a need for systems to monitor clinical performance of care processes and service delivery to reduce unnecessary morbidity and mortality. The aim was to use Shewhart’s statistical process control charts as a tool to monitor differences in stabilisation time for neonatal retrievals.

Methods We studied stabilisation time (time from first look till team is ready to depart referring hospital) retrospectively for immediate (patient in extremis or deteriorating) neonatal retrievals transferred by ground and air from 150 general and regional hospitals to 10 tertiary neonatal centres in NSW and ACT from 2000–2006. Shewhart’s charts were used based on standard deviation and mean rather than external “standard” criterion to detect aberrations in stabilisation time of transfers in a centralized transport service in New South Wales (NSW) and the Australian Capital Territory (ACT), Australia. “Out of control” signals were defined as stabilisation time above 2σ (warning) and 3σ (control) limits.

Results 4658 newborns were retrieved of which 780 (16.7%) were immediate. 37 of 780 (4.7%) were flagged as “out-of-control” signals for stabilisation time. Median (IQR) stabilisation time was 2.5 times longer for flagged cases compared to non-flagged; 348 mins (296–397) vs 132 mins (88–174). Reasons for “out-of-control” signals (>2σ) included severe disease (67.6%), junior registrar (45.9%) and equipment failure (18.9%).

Conclusions Shewhart’s control charts are a simple statistical tool to drive remedial changes. Although our study concerns monitoring aberrations in retrieval services, its approach and methodology is applicable across a wide range of specialties.